Re: Conlang Irregularities

Sally Caves wrote:

> Here's a new question that I'm curious about,
> and that I didn't ask on the Lunatic.:
>
> I know that one of the difficult tasks of getting
> a conlang up and running is to make the rules
> of grammar.  How many of you established
> conlangers, after having done that, deliberately
> introduce irregularities and contradictions into
> your conlang with an eye to giving it dimension
> and realism?  Or maybe you don't do it so
> deliberately... maybe it just happens and you
> decide to leave it be?
>
    I'm more of a  "just happens" conlanger. It seems to me that life
itself,  being full of irregularities and contradictions,
if a language is part of one's life for any great length of time, it is
bound to accumulate changes that can confuse and even contradict earlier
stages of the language. I tend to regard this as a sign of
_development_  rather than a sign of a problem.

     An example of "just happening" in rtemmu:
    According to the rules, a word describing something with gender is
made masculine by adding the derivational prefix  "mtu-" and made
feminine by adding the prefix "ela`-" (a` = open front unrounded vowel),
thus:
        vbr`x = dog         mtuvbr`x = male dog         ela`vbr`x =
female dog
but for  male and female _people_, the terms are " mtuku" = man,
"ela`kuh" = woman,  where [ku]and [kuh]  have no particular meaning.
Why not  "mtumpagi" or  "ela`mpagi?"  That's the way the language is.
(Actually, mtumpagi and ela`mpagi do exist. In order to account for the
difference, the rule has developed that they refer to strictly physical
descriptions, whereas mtuku and ela`kuh refer to the broader context of
manhood and womanhood, even though this confuses things since all other
uses of mtu- and ela`-  can refer to both physical and nonphysical
descriptions.

> In other words, how many exceptions to the rules
> you' ve made will you tolerate?  One of the
> criticisms leveled at invented languages is that
> they are too regular.  Does that bother you?
>
Dead, unchanging wood can make some pretty useful tools, but I like
live  (and complex) trees too.  :-)

> I have the choice of modifying my volitional verbs
> that end with an "n" and that have done so for twenty
> years.  _Euan_, for instance, means "to go," volitionally..
> But I've fairly recently made it a rule that non-volitional
> verbs will end in "n" in their absolute form (retaining the
> vestige of the gerundive suffix that marks them as
> non-volitional:  -ned.)  So:  teprorem, y tepro,  "touch,"
> "I touch"; but teproned, y tepron, "feel," "I feel."  Brilliant!
> But what do I do with euanrem y euan, "go," "I go," denrem
> y den, "tell, I tell," and uenrem, y uen, "take, I take"?
> Except to make them exceptions to the rule?  These
> are beloved words that I've had for decades and don't
> want to mess with.  And yet I can predict how they might
> change in subsequent usage: _euaned_ "go" but not with
> volition, and a new vol. form,  euarem, y eua--which I'm not
> crazy about.  And an epinthetic vowel inserted between
> root and suffix for uenrem and denrem:  _uenarem, denarem_.
> which are other verbs.  Yikes, what to do?
>
Maybe the new rule could continue to evolve until the conflict is
resolved.
If not, maybe it's not so bad tolerating some level of ambiguity in the
language. After all, aren't there natlangs which use the same affixes
for a number of  different grammatical tasks? Talk about confusion for
one trying to learn them! And yet, native speakers don't seem to mind.


Dan Sulani

--
 likehsna  rtem  zuv  tikuhnuh  auag  inuvuz  vaka'a.

 A  word  is  an  awesome  thing.